Transport of Sand Mixtures Including Consideration of Field Data, Threshold of Motion, Fai1 Velocity and Bed, Suspendedand Total Load Transport
Total Page:16
File Type:pdf, Size:1020Kb
fttOranlicsResearctt Wa[ingford TRANSPORTOF SANDMIXTURES A LITERATUREREVIEW April 1989 Report SR 200 Registered Office: Hydraulics Research Limited, Wallingford, Oxfordshire OXl0 8BA. Telephone: O49l 35381. Telex: 848552 This report describes work funded by the Department of the Environment under Research Contract PECD7/6/II2 for which the DoE nominated officer was Dr R P Thorogood. It is published on behalf of the Department of the Environment but any opinions expressed in this report are not necessarily those of the funding Department. The work was carried out by Dr N Walmsley and Dr G V Miles in the Tidal Engineering Department of Hydraulics Research, Wallingford, under the managementof Mr M F C Thorn. @Cro*n copyright 1989 Published by permission of the Controller of Her Majesty's Stationery Office ABSTRACT The study of sediment transport generally is very difficult but more so in the case of estuaries because: - the water movementsare continually changing with the rise and faII of the tide - certain sediments are not found in someparts, leading to unsaturated loads in the water - a wide range of sediment exist,s on the bed and in suspension. In recent years sediment transport models have been developed and used for rnaking engineering assessments of the impact of works on the sediment regime. At present the fulI potential of the models cannot be realised becauseof the lack of calibration and verification data, and gaps in our understanding of the fundamental sedimentation processes. Recent research, funded by the Department of the Environment under Research Contract PECD7/6/56, demonstratedthat computermodels of sediment transport can simulate the effects of variable tidal movementsand partly saturated loads of sediment. The objeetive of the present research project is to consider the problems associated with mixtures of sediment. The first phase of this project included a literature review of the transport of sand mixtures including consideration of field data, threshold of motion, faI1 velocity and bed, suspendedand total load transport. This report summarisesthe main references on the subject and contains a brief description of the latest theories being considered by other workers in this field. The main factors which will need to be taken into account in the computer model of sediment mixtures have been identified as the incorporation of different saturation concentrations for different mixtures of size fractions, different bed exchangerates resulting from different settling velocities and armouring of the bed by coarser grains, and variable bed composition arising from differential settling and resuspension. The information obtained from the literature review will be assessedin the next stage of the work and used to formulate a computer model for simulating the transport of mixtures. This will be described in the main contract report due in the Spring of 1990. CONTENTS Page 1 INTRODUCTION 1 2 GENERALBACKGROUND 2 2,I Mixture theory z a 2.2 Field and flume studies z 2.3 Numerical models 4 TRANSPORTOF SANDMIXTURES 5 3. I Threshold of sediment motion 5 3.2 Fa11 velocity 7 3.3 Bed load transport X 3.4 Suspendedload transport 10 3. 5 Total l-oad transoort L2 4 DISCUSSION T2 5 ACIO{OWLEDGEMENTS 13 6 REFERENCES I4 APPENDICES Appendix I - Main workers in the field Appendix II - Additional references 1 INTRODUCTION The numerical" model prediction of sediment transport plays an import.ant role in assessing the impact of engineering works in tidal est,uaries. The models can be used to assess deposition,/erosion problems during the early design stages and can identify possible future maintenance commitmenLs such as the dredging of shipping channels. A lot of work has been done in the past on the motion of uniform sands in one dimensional f1ows. In tidal estuaries the problems become more difficult because the water movements are continually changing with the rise and fall of the tide, a wide range of sediment exi-sts on the bed and in suspension and certain sediments are found in some parts but not in others which may lead to unsaturated loads in the water co]umn. Tn order to irnprove the usefulness of existing sediment transport models in engi-neering applications a number of developments have been outlined in relation to the modelling sand mixtures: - incorporation of different saturation concent.rations for different size fractions - different bed exchange rates resulting from variable settling veLocities - variable bed composition arising from differential erosion and deposition - the possibility of armouring of finer grains by coarser grains - efficient and accurate methods of calculating saturation concentrations - ini-tialisation of bed composition in situations of variable supply of erodible material The aim of this l-iterature review was to identifv the key workers in this field of study and to bring together many of the relevant references. This interim report is not an in depth technical review but is intended as a reference guide for the researchers undertaking the next phase of the research contract. GENERAL BACKGROUND 2.L Mixture theory The transport of fine material-s in turbulent fluid flow plays an important role in a broad range of disciplines and this has been refLected in the research literature. Fluid-solid mixture theory has been developed (McTigue, 1981) which models both the fluid and particulate phases as continua whilst taking account of turbulent fluctuations of the velocities and concentrations. Constitutive equations describing the shear stresses occurring in a mixture have been derived by Ackermann and Shen Q982) [see also Shen and Ackermann ( 1982) I whil-st other researchers have investigated bouyancy forces and concentration distributions in fluid-soIid mixtures (Zhaoyin, L987, and Zhaohui and Ti-ancheng, I9B7). 2.2 FieId and flume studi-es One of the drawbacks of modelling transport of sand mixtures in estuaries is the dearth of field data. However, in recent years a number of in-situ sedimentation studies have been undertaken by Vale and Sundby (1987), Jarvis and Riley (1987), Bedford et aI (1987) which reflect an increasing concern in the lack of relevant field data for the assessment of transport formul-ae and rnathematical model-s. One of the most extensive sedimentation surveys to be reported i-n the literature was carried out by Allen (1971) in the Gironde estuary. He monitored tidal current and lrave induced transport and investigated their effect on the grain size distributions in the estuary. This data was later used by Owens (1986) as validation data for his sediment transport model. The pattern of naturally occurring grain size distributions is often approximated by a Gaussian or normal- distribution where the 1og histogran approximates to a parabola. Bagnold and Barndorff-Nielsen (1980), however, found that measured sand grain distributions often approxi-mate to a hyperbolic distrj-bution. Their findings were further substantiated by Deigaard and Fredsoe ( 1978) who showed, by simulation, that originally log-norrnally distributed sand will tend to become J-og-hyperbolically distributed after sorting. Work on size distri-butions of sand has been eontinued by Barndorff-Nielsen and Christiansen (1988) whose field studies have shown that the mass-size distributions of sands vil1 vary depending on whether they are from a predominantly depositional environmental- or predominantly erosional environment. The development of numerical- models to simulate graded sediments has demanded that there be complementary physical model studies in order to gain a full understanding of the physical processes encountered in graded sediments (White, I972; Bridge, 1981; Westrich and Juraschek, 1985; Ribberink, 1987). In particular, Allard (1987) has carried out a series of well controlled flume experiments using sand particles in an effort to isolate certain phenomenaand to assess their relative importance. More recently, Wilcock and Southard ( 19BB) undertook a large range of flume studies on a range of graded sediment samples to investigate incipient motion of the individual grain sizes. Rakoczi (1987) has aLso investigated the effect of sediment mixtures on initiation and development of sediment motion in a series of Iaboratory investigations . 2.3 Numerical models The majority of sediment transport research, whether it be flume experiments, field studies or numerical model development, has been carried out for steady uniform flow conditions. Efforts have been made to develop more advanced models to handle unsteady and non-uniform conditions (Brownlie, 1981; Krishnappen, 1981; Tsujinoto, 1987 and Lyn, 1987) but these are generally only applicable to sediment transport in river channels. However, Markofsky et aI (1985) have developed a model specifically for use in tidal waters. Using their model, they investigated the sensitivity of model results to numerical as wel-l as physical parameters such as discharge, l-ocation of sediment source, stratification and tidal motion. In particular, they investigated methods of limiting the numerical- diffusion generated within the model. In recent years a number of one dimensi-on models have been developed to take account of graded sediments (Borah et aI, I9B2; Karim and Kennedy, 1985; Karim and Ho11y, 1986) which show good correlations with flume experiments but which falter vhen compared with field dat,a. Lu and Shen (1985) chose Lo investigate the use of an existing sediment transport formula, i.e Einstein's, and found that Einstein's hiding factor could in fact provide reasonable bed load esLimales. Ce1ik and Rodi (1985) proposed a model which involved new approaches for determining eddy-diffusivity, transport capacity and bed-boundary conditions. The model was applied to non-equilibrium transport situations and close agreement between prediction and measurement was found. An afternative two-dimensional vertical model was proposed by van Rijn (1986) in which finite-element methods lrere used to solve the convecti-on-diffusion equation. The model developed by Lee and Ahn (1986) r.rasderived from convective-diffusion theory but they included a hindered settling effect of the particles and the vertical veloci-ty component due to tidal movement.